Assignment 3: Bancassurance

 

Build a model to identify the positively responding customers who have a higher probability of purchasing the insurance

Contents:

  1. Data Overview and EDA
  2. Outlier Treatment and Feature Engineering
  3. Building Decision Tree Model and Analysis
  4. Building Logistic Regression Model and Analysis
  5. Conclusions

1. Data Overview and EDA:

General Observations:

Drop CUST_ID:

Fixing Types:

Explore Unique Values and Drop Duplicates:

2. Outlier Treatment and Feature Engineering:

Apply Log Scaling to Skewed data:

Outlier Treatment:

Comapare corrolation results:

One Hot Enconding and new Feature Engineering:

Feature Engineering:

Descrete Data Feature and Relationship to Product Purcahse:

3. Building Decision Tree Model and Analysis:

Decision Tree With Default Params:

Decision Tree With Grid Search:

4. Building Logistic Regression Model and Analysis:

5. Conclusions